Development and characterization of molecular glue degraders of CK1α and IKZF2 as therapeutic agents in oncology

The immunomodulatory imide drugs (IMiDs) thalidomide, pomalidomide, and lenalidomide are the foundation of a class of compounds called molecular glue degraders. These compounds regulate the recruitment of proteins to the E3 ligase substrate adapter cereblon (CRBN) within cells, which leads to the protein target’s ubiquitination and degradation. Designing molecules to recruit disease-relevant proteins holds significant therapeutic promise across many different disease indications, including oncology. Investigations into the possible range of proteins that may be recruited to CRBN have revealed common structural elements that enable recruitment to a CRBN-molecular glue complex. These structural elements are represented across the proteome and are contained within proteins with diverse cellular functions such as transcriptional regulation, apoptotic processes, cellular differentiation, cell cycle control, and angiogenesis, highlighting the broad therapeutic and scientific potential of this modality. Here, I describe the structure-activity relationship (SAR)-guided development of a novel CRBN-targeting molecular glue library to degrade casein kinase 1 alpha (CK1α) and transcription factor IKZF2 with application as a targeted therapy for acute myeloid leukemia (AML), gynecologic cancers, and breast cancer. In Chapter 1, I outline the history of the IMiDs and describe the canonical structural basis of their recruitment of proteins to CRBN. I discuss the general design principles of molecular glue degraders, including those specifically targeting CK1α and IKZF2. I review these proteins’ biological roles in pathways implicated in various cancers and summarize the current efforts to target CK1α and IKZF2 in the clinic. I also discuss recent advances in our understanding of the structural basis of the recruitment of proteins with non-canonical structural motifs to CRBN. In Chapter 2, I report the medicinal chemistry efforts to develop selective and potent molecular glue degraders of IKZF2 and CK1α with optimized pharmacokinetic (PK) properties and activity in AML models. I discuss the structural basis for these compounds’ ability to recruit IKZF2 and CK1α to CRBN and, starting with an early lead compound DEG-35, I describe the small library of analogues that I designed, synthesized, and evaluated to optimize DEG-35 for solubility and metabolic stability. These efforts resulted in DEG-77 which potently degrades IKZF2 and CK1α and has significantly improved metabolic stability, solubility, and in vivo PK properties. I report that DEG-35 and DEG-77 selectively degrade IKZF2 and CK1α across multiple AML cell lines and demonstrate antiproliferative effects by inducing apoptosis and differentiation. I then describe the validation of these molecules as molecular glues and show that IKZF2 degradation leads to differentiation while CK1α degradation leads to apoptosis, cell cycle arrest, and differentiation. Finally, I report the use of DEG-35 and DEG-77 in vivo in various AML models and show that treatment significantly delays AML disease progression, nearly doubling the lifespan of mice with AML. In Chapter 3, I describe our efforts to expand the therapeutic use of DEG-77 to solid tumors including p53 wild-type gynecologic cancers. I detail the biomarkers of sensitivity derived from our bioinformatic reanalysis of the characteristics of p53 wild-type cells sensitive to DEG-35 in the PRISM screen including cellular dependence on MDM4 and loss of function in the SWI/SNF chromatin remodeling complex. I then demonstrate the broader therapeutic utility of DEG-77 in subtypes of ovarian and endometrial cancers with these characteristics and show that DEG-77 promotes on-target cell cycle arrest, senescence, and apoptosis in these cell models. I provide evidence showing that DEG-77 modulates MDM4 protein levels and complexation with p53 likely via loss of MDM4 Ser289 phosphorylation upon CK1α degradation. Lastly, I demonstrate the efficacy of DEG-77 in promoting tumor growth inhibition in in vivo models of ovarian cancer and provide data motivating the rational combination of DEG-77 with existing clinical therapeutics. In Chapter 4, I explain the role of CK1α in the phosphorylation of Fas-Associated Death Domain protein (FADD), which allows breast cancer cells to escape cell cycle arrest mediated by CDK4 and CDK6 inhibition. I show that CK1α degradation by DEG-77 leads to potent cell cycle arrest and eventual senescence in a wild-type p53 breast cancer cell line, which is mediated by loss of FADD phosphorylation and upregulation of p21. I demonstrate that DEG-77 resensitizes cells to CDK4/6 inhibition and that their combination results in significant tumor growth inhibition in a mouse model of luminal ER+/HER2– breast cancer. In Chapter 5, I describe our investigation of how CK1α molecular glue degraders, including selected DEG compounds, affect protein levels of CK1α interactors Scaffold Anchor of CK1 (SACK1) proteins A–H. I report that select SACK1 proteins are degraded with various CK1α degraders and demonstrate that SACK1G co-degradation in the presence of a CK1α degrader requires interaction with CK1α, supporting a “bystander degradation” mechanism wherein the co-recruitment of SACK1 proteins to CRBN when in complex with CK1α leads to their co-ubiquitination and degradation. Finally, in Chapter 6, I outline the possible path to the clinic for DEG-77 or a similar dual-targeted IKZF2 and CK1α molecular glue degrader. I also expand on future experiments to further probe the role of MDM4 and loss of SWI/SNF function in mediating cellular sensitivity to CK1α degradation. Lastly, I describe the broader potential clinical application of a CK1α and IKZF2 dual-targeted degrader in malignancies driven by biological pathways outside of pFADD and p53 in which CK1α plays a critical role.